Basic ROOT Use

Today we’ll be taking this raw data and plotting a histogram in ROOT.

Histograms

A histogram is a graph of a set of observations. Say we have n observations of a random variable x_i, called bins. The number of occurences of n_i of x in each subinterval is plotted on the y-axis. In the limit of zero bin width, and infinite observations a histogram exactly replicates a probability density function, or p.d.f. This shows the relation between histograms as we change the number of observations and the p.d.f. from which observations are drawn.

To create a histogram from input data text file Day1/basic.dat simply run the macro named Day1/histogram1D.C. To run it, begin at a UNIX prompt in the directory Day1, open ROOT, then execute the macro

  $ root -l
  root [0] .x histogram1D.C
  x=-1.102279, y=-1.799389, z=4.452822
  x=1.867178, y=-0.596622, z=3.842313
  x=-0.524181, y=1.868521, z=3.766139
  x=-0.380611, y=0.969128, z=1.084074
  x=0.552454, y=-0.212309, z=0.350281
  root [1]

Tree

A ROOT TTree (just “tree” for short) is structured container of data, specially designed for collider. In a collider, we typically are interested in the distribution of certain quantities over many collision events (or entries). For example, in events with electrons, we may be interested in the distribution of the an electron’s momentum |p|, maybe after requiring that the electron was within “acceptance” of the detector.

A tree tries to organize this information by effectively linking similar types of information from different events. This link is called a branch. In the tutorial I showed how to create a tree, with a few branches for the variables x, y, and z. Sometimes this is called a “flat ntuple” because none of the information is nested. Here’s a more complicated example, with a nested hierarchy of information. There is a branch of particles, each particle branch has sub-branches: http://lcg-heppkg.web.cern.ch/lcg-heppkg/ROOT/eventdata.root